Many mammalian genes are too large to be cloned intact in standard Lambda bacteriophage or cosmid systems. Furthermore, large DNA fragments (greater than 100 kilobases) are particularly difficult to manipulate as they are prone to breakage by cleaving during handling. However, in order to study, at the DNA level, some metabolic diseases, the regulation of metabolic pathways, or the coordinate regulation of any group of genes, it may be necessary to transfer, from cell to cell, DNA fragments spanning as much as several million base pairs. This proposal involves the development of techniques to transfer very large DNA molecules from Saccharomyces cerevisiae cells to mammalian cells, and vice versa, without direct biochemical manipulation of the DNA. This will be done by determining cell fusion protocols which allow selection of either yeast cells which have received mammalian DNA, or mammalian cells which have received DNA from yeast. These procedures will enable us to construct libraries of mammalian DNA in yeast, and to screen these libraries for specific mammalian genes by their fusion with appropriate mammalian cell mutants. Also, we will be able to transfer particular large genes or DNA sequences from one genetic background to another for experimental analysis. These experiments would seem essential for devising successful gene therapy protocols for treatment of inborn errors of metabolism. Together with pulsed field gel electrophoresis, these methods will greatly expand our ability to analyze mammalian genomes.